PulseAugur
EN
LIVE 14:07:23

Anthropic's Numbat model checkpoint leaked online

By PulseAugur Editorial · [1 sources] ·

A leaked checkpoint for Anthropic's unreleased "Mythos-class" model, codenamed Numbat, has surfaced online. The checkpoint, which is approximately 7.72 TB and contains 142 tensors, is described as a mid-pretraining artifact from a large-scale training run. While it includes a reference implementation for loading and running the model, it is raw pretraining output without any post-training or RLHF, making it unsuitable for direct inference. AI

IMPACT Leaked model checkpoints can offer insights into frontier model architectures and training techniques, potentially accelerating research.

RANK_REASON This is a leaked checkpoint of an unreleased model, not an official release from the lab.

Read on HN — anthropic stories →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Anthropic's Numbat model checkpoint leaked online

How we ranked this

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
This is a leaked checkpoint of an unreleased model, not an official release from the lab.
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
1 days old
Coverage has settled into its steady-state source set.

Full methodology in our editorial standards.

COVERAGE [1]

  1. HN — anthropic stories TIER_1 English(EN) · explosion-s ·

    Alleged Mythos Class Anthropic Checkpoint Leak

RELATED ENTITIES

RELATED TOPICS